It came from outer space

Hollywood loves its aliens, but how plausible are they? Seth Shostak (left), an astronomer with SETI – the institute that seeks out extra-terrestrial life – has a good idea of what might really be out there. Steve Connor meets him

Seth Shostak has a vision of aliens from outer space. It is not one of little green men or bald, bulbous-brained geeks. His is a rational vision, one built on hard facts and intelligent speculation, which is hardly surprising given that he is senior astronomer with the SETI Institute in California. Forget what you've seen on The X-Files, SETI – the search for extra-terrestrial intelligence – is the serious, scientifically based endeavour of looking for signs of other civilisations beyond Earth.

Perhaps not surprisingly, Shostak has good reasons to believe that ET is out there, and that we have a fair chance of finding him within the next 20 or 30 years. If we ever do pick up the radio signals of intelligent aliens, it will almost certainly be one of the most momentous events in human history. No doubt it will lead to even greater speculation as to what the real ET might look like. Shostak, who was in Britain last week to open the SETI theatre show at the National Space Centre in Leicester, has been pondering this alien conundrum.

"Of course we don't know what they look like. Only Hollywood knows for sure," says Shostak, who wrote the script for the show. "According to Hollywood, aliens are pretty anthropomorphic – they look a bit like us. But that's understandable, because if they look like soldering irons or vacuum cleaners we wouldn't be able to identify with them."

Aliens may not be likely to look much like us, but there are still some basic laws of physics governing their overall physiognomy. "The human form has a lot of features that make sense in an engineering way," says Shostak, as he brings his forefinger and thumb to his eye as if inspecting a speck of dust. "You know that ET is not going to be this big because there seems to be a minimum size for neurons. If you make ET too small, he's not going to be terribly bright."

Neither is he going to be much bigger than, say, an elephant. "You can probably rule out ET being too large because his planet is going to be roughly the same size as ours."

And if ET lives on a planet the size of Earth, this means living under a similar force of gravity, with a similar set of physical constraints on overall body size. "If you were to double your size, that would multiply by four the cross-sectional area of your muscles, meaning that you would be four times as strong," Shostak explains. "But you'd be eight times as heavy because you'd be twice as high, twice as wide and twice as deep. So your power-to-weight ratio would go down by a factor of two. Being bigger makes you less agile and less strong, which is why ants can lift many times their body weight."

So not too big, not too small, and all because ET is likely to live on a planet roughly the size of Earth. And the reason why he (or she, or it) does that is because a rocky planet that's the size of Earth, with an atmosphere as well as a decent force of gravity, is the best environment for keeping liquid water – that vital ingredient of life – from just floating off into space. "We think it'll be a planet with liquid water but a little bit of land. It it's all water and oceans, it's unclear whether you'll ever develop very much intelligence," he says.

"Most of the things in the ocean are not terribly bright. Lobsters don't hold up their side of the conversation very well. If you look at what is bright in the oceans, it tends to be things that once lived on land, like whales and dolphins."

Shostak suggests that alien intelligence would only get a real chance to develop its full potential if it evolved on land. "The argument that is often made is that the ocean is too easy an environment. The weather under the ocean doesn't change, so there is rather little competitive stimulus and so forth to lead to the development of great intelligence."

So, we have an ET that is not too big, not too small, and a land-lubber with a thirst for water. What else can we imagine them to look like? Again, we can call upon some astronomical details of ET's home planet to work out his evolutionary identity. "It turns out that stars like the Sun put out the bulk of their energy in the wavelengths of light that we see," says Shostak.

So, ET will need something to see with, and an atmosphere means that he might want to hear what he cannot see. "ET will probably have eyes and ears, which are useful when your sightline is short. You may not always see everything, but you can hear it, so there is some advantage in having ears.

"How many eyes? Well, one eye is a big improvement over no eyes, but two eyes is a big improvement over one eye because of stereo vision. If you need to catch dinner, stereo vision helps," he says. "Three eyes might be good. You might imagine having an eye in the back of your head to see what's chasing you, but there are not many animals that have done this on Earth, which suggests that the extra costs incurred are not worth the additional benefit. ET probably won't have 100 eyes. It will need too much brain power to handle all that visual information."

Our identikit image of ET is taking shape. An average-sized land-dweller with a couple of eyes and ears, who cannot move very far away from water. It's beginning to sound familiar, but then Shostak changes tack. "All this discussion may be appropriate in terms of telling us how ET might have evolved, but it may not tell us much about the critter that we actually find," he says, rather mysteriously.

The fact of the matter is that even if we detect radio signals of a distant civilisation, it will still be too distant to realistically contemplate a close encounter. "It's very difficult for us to reach the stars. It takes 100,000 years to go to the nearest at the speed of current rockets. If we can build rockets that are 100 times faster, it's still 1,000 years away," he says.

But if an alien civilisation was further advanced than us, it is likely to have cracked the problem of machine intelligence. If so, it would be far more plausible that our first encounter would be with the alien's machines than with the aliens themselves. "Artificial intelligence can improve itself very quickly. The first thing you'd do with a thinking machine is to have it design something that is better than itself. It's not Darwinian evolution, it's directed evolution. At that point they would take over their planet – and not only would they be in charge, they would be able to travel."

If we are ever to come across ET, then, it is more likely that we'll meet his invention: a machine with an intelligence to outwit our own. If so, the psychology of such an encounter may be somewhat different to that which would involve two sets of intelligent organisms with a finite lifespan. Machines can be immortal.

"If you were immortal, you'd start worrying about very long-term dangers. If these machines are in that category, they would be worried about really big-picture things, like the end of the universe," says Shostak. "They would probably spend a lot of time doing what Stephen Hawking does – trying to figure out if there is a parallel universe that they can get into, because in the end, they've got to get out of this one."

So, in the unlikely event that our corner of the Milky Way will ever get a visit from ET's intelligent, robotic nemesis, we can at least rest assured that it is likely to have other things on its mind than the mundane fears of an ape-like creature living on a tiny blue dot of a planet. Unless, of course, Stephen Hawking, or someone of a similar intellectual stature, suddenly finds the key to a parallel universe – and this is the key that the alien machine wants.